Quantitative mRNA Imaging with Dual Channel qFIT Probes to Monitor

Jan 29, 2018 - Monitor Distribution and Degree of Hybridization. Imre Gaspar,. †. Felix Hövelmann,. ‡. Jasmine Chamiolo,. ‡. Anne Ephrussi,. â€...
0 downloads 0 Views 3MB Size
Articles Cite This: ACS Chem. Biol. 2018, 13, 742−749

Quantitative mRNA Imaging with Dual Channel qFIT Probes to Monitor Distribution and Degree of Hybridization Imre Gaspar,† Felix Hövelmann,‡ Jasmine Chamiolo,‡ Anne Ephrussi,† and Oliver Seitz*,‡ †

European Molecular Biology Laboratory (EMBL) Heidelberg, 69117 Heidelberg, Germany Institut für Chemie der Humboldt-Universität zu Berlin, 12489 Berlin, Germany



S Supporting Information *

ABSTRACT: Fluorogenic oligonucleotide probes facilitate the detection and localization of RNA targets within cells. However, quantitative measurements of mRNA abundance are difficult when fluorescence signaling is based on intensity changes because a high concentration of unbound probes cannot be distinguished from a low concentration of targetbound probes. Here, we introduce qFIT (quantitative forced intercalation) probes that allow the detection both of probe− target complexes and of unbound probes on separate, independent channels. A surrogate nucleobase based on thiazole orange (TO) probes the hybridization status. The second channel involves a nonresponsive near-IR dye, which serves as a reporter of concentration. We show that the undesirable perturbation of the hybridization reporter TO is avoided when the near-IR dye Cy7 is connected by means of short triazole linkages in an ≥18 nucleotides distance. We used the qFIT probes to localize and quantify oskar mRNA in fixed egg chambers of wild-type and mutant Drosophila melanogaster by wash-free fluorescence in situ hybridization. The measurements revealed a relative 400-fold enrichment of oskar within a 3000 μm3 large volume at the posterior pole of stage 8−9 oocytes, which peaked at a remarkably high 1.8 μM local concentration inside 0.075 μm3 volume units. We discuss detection limits and show that the number of oskar mRNA molecules per oocyte is independent of the oocyte size, which suggests that the final levels are attained already during the onset of oskar localization at stage 8.

T

For imaging of mRNA in nonmodified wild-type cells, in situ hybridization with fluorescent oligonucleotide probes usually is the method of choice. Single-molecule fluorescence in situ hybridization (smFISH) is a powerful method that involves the use of tens of labeled oligonucleotide probes designed to bind to various regions of the mRNA target.13 Prior to analysis of fixed cells, unbound probes must be removed by washing. Because each mRNA target is loaded with tens of fluorophores, low abundance targets can be quantified by means of counting of single molecules. The use of the hybridization chain reaction provides single molecule sensitivity in mRNA imaging with a set of two oligonucleotides.14,15 Fluorogenic hybridization probes have been developed with the aim to speed up analysis times by enabling wash-free FISH.16−34 In addition, fluorogenic hybridization probes allow RNA imaging of live cells where separation of unbound from bound probes is impossible. Probes such as molecular beacons,18−20 FIT probes,21−24 binary probes,25−27 ECHO probes,28,29 reactive probes,30−34 and others16,17 exhibit enhancements of fluorescence upon sequence-specific hybridization owing to changes in (i) the

he knowledge of gene expression levels is essential for our understanding of cellular processes. Numerous methods such as sequencing,1 Northern blotting,2 and quantitative PCR3,4 give insight into the RNA expression of an average of the population of the cells investigated. When spatial and temporal resolution in single cells or in tissue is needed, more advanced techniques are required. The ideal method provides quantitative information about localization and copy numbers of the mRNA target of interest. Frequently, the mRNA of interest is genetically tagged with repeats of recognition units (e.g., MS25 or λN226 hairpins) that recruit coexpressed autofluorescent fusion proteins. Because each mRNA target is loaded with tens of fluorophores, low abundance targets can be quantified by means of counting of single molecules. In a somewhat related approach, RNA aptamers that fluoresce upon binding of profluorescent dyes have been used for staining of mRNA in bacterial cells and miRNA in eukaryotic cells; however, applying the method for imaging eukaryotic mRNA has proven challenging.7−10 Autofluorescent protein fusions with RNA-binding proteins based on PUMILIO proteins11 or the CRISPR system12 avoid the need for RNA-based tag sequences. Though successful in many instances, the approaches call for genetic modification of cells, and background from unbound fluorescent protein fusions can be high. © 2018 American Chemical Society

Received: November 27, 2017 Accepted: January 29, 2018 Published: January 29, 2018 742

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology distance between a fluorophore and fluorescence quencher or (ii) the environment of locally responsive fluorophores. Fluorescence-activating hybridization probes face a common challenge: differences in efficiency of delivery, in subcellular localization and nuclear sequestration, and false positive signaling can have a dramatic effect on the fluorescence background in areas of low target concentration.22,35 This poses a challenge to quantitative measurements because any fluorescence signal can be caused either by a low number of target bound probes or by a high number of unbound probes. It is for this reason that the existing methods for counting RNA copy numbers typically involve wash steps or require repeat tags to increase contrast. We assumed that quantitative measurements of mRNA concentration should be feasible when self-reporting probes allow not only the detection of probe−target complexes, but also of unbound probe on a separate, independent channel. Such a multichannel system has been introduced by Tsourkas and co-workers, who used 65-nucleotide-long ratiometric bimolecular beacons (RBMBs) involving an extended, siRNAlike, double-stranded sequence and three labels to quantitatively assess the expression of engineered mRNA.36−38 Quantitative imaging of wildtype cells has not been demonstrated. We developed the so-called forced intercalation (FIT) probes that make use of a single, environmentally sensitive fluorophore of the thiazole orange (TO) family of dyes (Figure 1A).23,39 We have demonstrated that FIT probes

of unbound probes from a low concentration of target-bound probes. In this study, we introduce DNA qFIT-probes for quantitative RNA imaging (Figure 1B). The probes combine an environmentally sensitive TO nucleotide with a second, near-IR dye in a way that prevents fluorescence responses of the near-IR dye and leaves the brightness of the TO nucleotide unaffected. While the TO nucleotide serves as a reporter of hybridization, the nonresponsive near-IR dye provides information about probe concentration. To demonstrate quantitative mRNA imaging in complete tissue, we studied localization and determined the concentration of oskar mRNA in specific compartments of Drosophila melanogaster oocytes.



RESULTS AND DISCUSSION Design and Synthesis. In an ideal scenario, the two reporter dyes in qFIT probes should function as independent modules. The hybridization reporter should be dark in the absence of a target and emit bright fluorescence upon binding of the RNA target. It is therefore desirable to minimize quenching of TO emission by the concentration reporter. This calls for a near-IR dye, which minimizes spectral overlap between its absorbance and the emission of the TO dye. We furthermore sought for a means to prevent ground state interactions between the concentration reporter and TO. Unintentionally, this could make the fluorescence properties of the concentration reporter susceptible to changes of the probe’s hybridization status. We envisioned that a short tether and a net positively charged concentration reporter would help avoid contact with the positively charged TO dye and based on the ease of synthesis and functionalization chose the cyanine-7 (Cy7) dye (Figure 2A). The NIR dye was introduced to DNA FIT probes by means of alkyne−azide click chemistry (Figure 2B).47,48 Oligonucleotide synthesis was carried out according to previously reported methods.22,41 A serinol-TO phosphoramidite24 was introduced at three different positions (nt 6, 12, and 18) of a 24mer DNA-sequence which targets oskar-RNA in Drosophila melanogaster (Bp. 83−106 of CDS). The same

Figure 1. (A) FIT probes signal binding of complementary RNA by enhancement of fluorescence emitted from a thiazole orange (TO) nucleotide. (B) qFIT probes provide information about concentration and hybridization status by means of two detection channels, which include the environmentally sensitive TO dye and a hybridization independent near-infrared dye such as Cy7.

allow qualitative imaging of RNA in nonengineered cells.40−44 In FIT probes, the dye substitutes for a canonical nucleobase. The fluorescence emission remains low in absence of a target because twisting around the methine bridge rapidly depletes the TO excited state.45 Twisting is hindered in the target-bound form, when hybridization enforces intercalation of the TO dye. As a result, FIT probes furnish marked increases of fluorescence upon hybridization. Owing to the requirement for intercalation, FIT probes have low vulnerability to false positive signaling by nuclease degradation or protein binding.22 We reported methods to improve upon the hybridization-induced fluorescence enhancement and the brightness in the target-bound state.41,42,46 However, with signaling based on intensity changes, it was impossible to distinguish a high concentration

Figure 2. (A) Building blocks used for the synthesis of qFIT probes. (B) Cy7 is introduced via Cu-based click reaction with Cy7-N3 (1). Conditions: (a) 100 μM Bu-FIT DNA, 500 μM 1, 7.5 mM sodium ascorbate, 3.75 mM THPTA, 0.75 mM copper(II)sulfate in buffer (100 mM TRIS, 100 mM NaOAc, 1 mM MgCl2, pH 8), 3 h, 55 °C, 40−60% overall yield. 743

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology

Table 1. Sequences, Melting Temperatures, and Fluorescence Properties of Oskar Probes with and without Cy7 Labelling name

sequence 5′ → 3′

TM/°C

I0a

Ia

I/I0

I/I0 (Cy7)

osk1a osk1b osk1c osk1a_Cy7 osk1b_Cy7 osk1c_Cy7 (osk1q) osk2q osk5q

Bu-GAC TTX AGA TAA TAG GTT TTG GCG Bu-GAC TTA AGA TAX TAG GTT TTG GCG Bu-GAC TTA AGA TAA TAG GTX TTG GCG Cy7-Bu-GAC TTX AGA TAA TAG GTT TTG GCG Cy7-Bu-GAC TTA AGA TAX TAG GTT TTG GCG Cy7-Bu-GAC TTA AGA TAA TAG GTX TTG GCG Cy7-Bu-AAA AGC GGA AAA GXT TGA AGA Cy7-Bu-CGG TTT TCT GGX TTT GGG T

56 55 55 57 56 56 48 60

24 20 37 5 5 11 8 4

82 182 173 17 83 166 107 66

3.5 9.3 4.7 3.1 15.4 15.2 13.4 16.5

1.0 1.1 1.1 1.1 1.3

The experimental error due to inaccuracies of concentration determination is in the 10% range. Conditions: 0.5 μM probe and 5 equiv RNA-target, when added, in PBS (100 mM NaCl, 10 mM Na2HPO4, pH 7) at 25 °C. Red marked X denotes a Ser(TO) nucleotide (see Figure 1). TO: λex = 485 nm, λem = 535 nm; Cy7: λex = 690 nm, λem = 774 nm. a

osk1c_Cy7, where Cy7 and TO were separated by an 18 nt distance. Importantly, the readout of the Cy7-channel revealed independent emission (I/I0 = 1.0−1.3), regardless of the presence or absence of target RNA, and the brightness of the TO signal remained unaffected as long as the Cy7−TO distance was sufficiently large to avoid FRET in the doublestranded state. As an added advantage, the feasibility of FRET in the single stranded state increases the hybridization-induced response of TO emission. For application in fluorescence microscopy, qFIT probes must tolerate a biological matrix without a loss of performance. We evaluated the probes in lysate from 106−107 MDCK cells. The turbidity of the mixture that contained all cellular compounds such as membrane components, proteins, fragments of organelles, and nucleic acids showed a relatively high background signal. The responsiveness of TO and the Cy7emission was determined upon stepwise addition of the target (0.1 equiv steps) in order to simulate the absence of a target (0.0 equiv target), partial hybridization (0.1−0.9 equiv target), and full hybridization (≥1.0 equiv target). The TO emission increased until a 1:1 probe/target stoichiometry was reached (Figure 4A). Measurements in PBS buffer revealed that the TO quantum yield of osk-1c_Cy7 (ϕ0 = 0.05, ϕ = 0.29) was increased by a factor of 6 upon RNA binding. By contrast, Cy7 quantum yields remained unchanged (ϕ0 = ϕ = 0.29), and the independence of Cy7 emission persisted also in cell lysate over the complete range (0.0 to 1.6 equiv) of titration (Figure 4B). This should allow the calculation of a contrast-factor I(Cy7)/ I(TO), which indicates the degree of probe hybridization and could be used as a calibration for application in cells. Quantitative mRNA Imaging in Drosophila Oocytes. Previously, we used a combination of three fluorogenic hybridization probes (osk1, osk2, and osk5) for targeting different sites of oskar mRNA in Drosophila oocytes (see Figure S7).41,42 This procedure increases the sensitivity of RNA imaging (vide infra). In addition to osk1q (identical to osk1c_Cy7), we converted the osk2 and osk5 FIT probes to qFIT probes osk2q and osk5q, respectively (Table 1). The probes showed, again, hybridization dependent TO emission, while Cy7 emission remained virtually unaffected by hybridization (e.g., osk5q: ϕ0(TO) = 0.06, ϕ(TO) = 0.27, ϕ0(Cy7) = 0.22, ϕ(Cy7) = 0.23). We performed a rapid, wash-free FISH employing osk1q or a combination of osk1q, osk2q, and osk5q in fixed, stage 8−9 Drosophila oocytes, in which oskar mRNA begins to accumulate in great quantities at the posterior pole of the oocyte.42 Confocal fluorescence microscopy was used to obtain the signal of Cy7 (Figure 5A) and TO (Figure 5B) dyes of the qFIT probes, which report the local concentration and

segment was previously targeted with a molecular beacon by Tyagi et al. for in vivo studies.49 The alkyne was introduced at the 5′ end by using butyne-1-ol-phosphoramidite 2 for coppercatalyzed click reaction. The conjugation of Cy7 azide 1 proceeded smoothly in nearly quantitative yield by applying a procedure developed by Brown et al. (see Figures S2 and S3 for HPLC-traces).50 Fluorescence with Synthetic RNA. A comparison of fluorescence spectra obtained in the absence and presence of a synthetic RNA target showed that, upon hybridization, all tested FIT probes provide enhancements of TO emission at 535 nm (Table 1). UV melt analyses revealed little, if any, differences in the thermal stability of the probe−target duplexes. We concluded that any change in fluorescence properties would solely be due to the type of Cy7 attachment and the Cy7−TO distance rather than by differential hybridization properties. The fluorescence response I/I0 furnished by qFIT probes was higher than the response by TO-only FIT probes provided that the TO−Cy7 distance exceeded 6 nucleotides (Table 1, see Figure S4 for UV and fluorescence spectra of all probes). For example, the TO-only probe osk1c showed a rather modest 5-fold intensification of fluorescence (Figure 3A), which improved to 15-fold upon

Figure 3. Fluorescence emission spectra of (A) osk1c and (B) osk1c_Cy7. Excitation of TO (λex = 485 nm, λem = 500−900 nm) before (black) and after (red) addition of target. Excitation of Cy7 (λex = 690 nm, λem = 700−900 nm) before (gray) and after (blue) addition of target. Conditions: 0.5 μM probe and 5 equiv RNA target, when added, in PBS (100 mM NaCl, 10 mM Na2HPO4, pH 7) at 25 °C.

introduction of the 5′-terminal Cy7 in osk1c_Cy7 (Figure 3B). The emission spectra revealed Förster energy transfer from TO to Cy7 (emission 750−850 nm) in the single stranded state, whichdespite the very small spectral overlapprobably is feasible, owing to the very short distance between TO and Cy7 in the unbound state. Of note, energy transfer and, thus, TO quenching was negligible in the target-bound qFIT probe 744

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology

Figure 4. (A) TO emission and (B) Cy7 emission spectra of osk-1c_Cy7 upon stepwise (0.1 equiv) addition of target RNA. Conditions: 200 pmol probe in MDCK lysate (106−107 cells/mL), target RNA added in 0.1 equiv steps (20 pmol). TO: excitation at 485 nm, integral of emission signal 500−700 nm. Cy7: excitation at 690 nm integral of emission signal 700−900 nm. All values are background corrected.

reasonable agreement with the values obtained by the qFIT probes. Oocytes that lacked half the amount of oskar mRNA (oskar0/+, see also Figure S10C) showed a significantly lower copy number of oskar mRNA ∼ 3.5 × 105 (1.4 × 105 to 5.9 × 105, p < 0.01 against both wild-type estimates). In the germ-line compartment of egg chambers lacking oskar mRNA completely (oskar null, see also Figure S10D), we estimated a negative number (−4 × 104 copies, Figure 5F), indicating that the method slightly overcorrects the raw measurements. We applied our method to correlate the estimated oskar mRNA copy numbers with the size of the oocyte during stage 8−9 (Figure 5G). The size of the analyzed wildtype oocytes varied between 0.48x × 105 to 3.78x × 105 μm3. Of note, there was no apparent correlation; small oocytes contained approximately the same number of oskar mRNA as large oocytes. A similar behavior was observed for the oskar0/+ oocytes. Thus, we conclude that the total concentration of oskar mRNA should decrease as the oocyte grows from stage 8 onward. Indeed, we found a moderate, negative correlation between the size of the oocyte and the total concentration of oskar mRNA, which reached a value below 10 nM by the end of stage 9 (Figure S12A). Previously, we defined the posterior domain (PD) in which oskar mRNA localizes as the posteriormost ∼9 μm of the anteroposterior axis.52 The local concentration of oskar mRNA within the boundaries of the PD (Figure 5A) increased continuously to above 300 nM as a function of oocyte volume (Figure S12B, 208 ± 80, n = 8). Given the constant number of oskar mRNA molecules in the oocyte, the local accumulation of oskar resulted in an exponential decrease in oskar concentration throughout the rest of the growing oocyte, from 16.5 nM to 0.8 nM (Figure S12C, 5.5 ± 4.8 nM, n = 8). Consequently, the relative enrichment of oskar mRNA at the posterior pole rose from around 10-fold to above 400-fold during the growth of the oocyte (Figure S12D). Within the PD, we previously found that 52 ± 18% (mean ± SD) of oskar mRNA localizes during stage 9 of oogenesis.52 Using our method described here, we estimated oskar mRNA copy numbers within the PD in oocytes already showing accumulation at the posterior pole (Figure 5F, red dots). In agreement with our previous measurements, we found 55 ± 20% and 49 ± 20% of the total mRNA within the PD of wildtype and oskar0/+ oocytes, respectively.

the hybridization status of the probe, respectively. In any given voxel, we assumed that the local concentration (cx,y) is a linear combination of the single-stranded and double-stranded probe species. By measuring Cy7 and TO emission and knowing the fluorescence of TO and Cy7 in a cell-free background as well as the responsiveness of emission of the given qFIT probes in a microscope setting (Figure S6), we determined the concentration of the probe-bound mRNA target (see SI for details). Figure 5E shows the distribution of the concentration of double stranded qFIT probes. The enrichment of oskar mRNA at the posterior pole of stage 8−9 Drosophila oocytes (Figure 5A and B, green dashed line) has frequently been observed.49 Our measurements (Figure 5E) revealed a surprisingly high 208 ± 80 nM concentration of oskar mRNA (mean ± SD, n = 8, combined results of osk1q and osk1q, -2q, -5q) within a 3140 ± 410 μm3 volume at the posterior pole (∼2.4% of total oocyte volume). Of note, we detected up to 1.8 μM local concentration per voxel in this posterior region, which exceeds the concentration of probes applied (0.5 μM osk1q, osk2q, and osk5q each, see also Figure S11). We considered the feasibility of a local enrichment of probes due to a concentration gradient induced upon binding of the target. Indeed, the concentration distribution measured by the Cy7 signal indicates the enrichment of probes in the posterior domain and depletion in the remaining regions of the oocyte (Figure 5A and C). For the estimation of oskar mRNA copy number in the oocyte, we introduced a threshold value which was set to the 97.5th percentile of the cds,x,y distribution measured in follicle cells (where oskar is not expressed): voxels having less than 97.5% of the maximum of the already low values found in this area were excluded from the downstream analysis (Figure S13A). To estimate the number of oskar mRNA molecules found within the oocyte, the raw copy number (Nraw) was calculated by using Avogadro’s law (see SI for details). The residual low intensity from TO remaining in the follicle cells after thresholding was used to estimate the mean false-positive detection rate which was applied to correct the raw copy number. Using osk1q probes and an equimolar mixture of osk1q, -2q, and -5q probes on wildtype oocytes, we obtained around 6.6 × 105 (3.9 × 105 to 1.1 × 106) and around 1.0 × 106 (5.9 × 105 to 1.6 × 106, p = 0.0513) copies of oskar mRNA per oocyte (Figure 5F). In recent experiments, oskar mRNA copy number was assessed by means of qRT-PCR. Values around 3 × 105 to 1.2 × 106 in Drosophila embryos51 are within 745

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology

Figure 5. Measuring concentration of oskar mRNA in developing Drosophila oocytes using qFIT probes. Maximal intensity projections of nonoverlapping optical slices of the Cy7 (A) and TO (B) signal of osk1q, -2q, -5q probe mixture (0.5 μM each). Dashed circles (A and B) indicate cell-free background. Asterisk denotes part of another oocyte (A); white arrow indicates autofluorescence of chitin in a trachea (B). The oocyte (white dashed line) and the follicle cell compartments (yellow dashed line) as well as the anterior margin of the posterior domain (PD, green dotted line) are indicated. (C) Local concentration of qFIT probes as determined by Cy7 fluorescence. (D and E) Local concentration of single (css, D) and double stranded (cds, E) qFIT probe species. Key indicates measured concentration values. (F) Corrected copy number of oskar mRNA estimated within the oocyte (black dots) and in the PD (red dots; see also Figure S13B). (G) Correlation between oskar mRNA copy numbers and the oocyte size. R2 values of the different experiments are indicated. Dot size indicates the logarithm of the oocyte volume. (H) Limit of detection per voxel (0.075 μm3) as a function of the threshold level as percentile of the cds distribution in the follicle cells. False positive detection rate (FPDR) associated with the threshold is indicated above the graph. Scale bars represent 50 μm (D and E).

alone (Figure 5H). However, based on the ratio between TO signals in the follicle cells after thresholding to the 97.5th percentile and TO in the oocyte, the false positive detection rate (FPDR) was estimated to be around 25%. As a result, only three-fourths of the voxels in the oocyte contain valid information on the mRNA concentration. By raising the threshold level to the 99.9th and 99.99th percentiles, FPDR

To assess the scope of mRNA quantification by qFIT probes, we determined the influence of thresholding on the lower limit of copy number per voxel. At the initially chosen 97.5th percentile of the cds,x,y,z distribution in follicle cells, we determined a lower limit of 3.3 ± 0.6 (mean ± SEM) copies per voxel (0.075 μm−3) when we used the equimolar mixture of osk1q, osk2q, and osk5q and 5.1 ± 0.7 copies by using osk1q 746

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology

use of a combination of three probes reduced the detection limit to three or nine copies per 0.075 μm−3 voxel at 25% and 0% FPDR, respectively, and we assume that single molecule specificity could be achievable with less than 10 qFIT probes. In this study, qFIT probes were used to analyze the localization and concentration of an mRNA target that is highly expressed within a large cell (1.0 × 106 copies within (0.5−3.8) × 105 μm3). It should be straightforward to apply the method for the quantitative analysis of mRNA localization in smaller sized cells. Given the estimated detection limits, we assume that the method in its current form will allow the facile quantification of mRNA targets expressed at medium−high abundance, when mRNA counting by smFISH loses accuracy owing to the need for fitting of data by multiple Gaussian functions. As an added advantage, qFIT probes allow quantitative mRNA imaging by wash-free FISH, which on the one hand provides for calibration by analysis of cell-free background and on the other hand avoids the time-consuming optimization of wash protocols.

decreased below 5% and 1%, respectively. At these confidence levels, the detection threshold was measured to be 8.0 ± 1.0 and 9.4 ± 1.1 copy numbers per voxel when using only the osk1q. We found a substantial reduction in these values to 5.8 ± 0.7 and 7.6 ± 0.9 copy numbers per voxel when we used the equimolar mixture of osk1q, osk2q, and osk5q (Figure 5H). Once the local concentration measured within the oocyte exceeds the maximum value observed in the follicle cell compartment, absolute confidence is reached at 12.3 ± 1.7 and 9.6 ± 1.5 copy numbers per voxel with the osk1q and osk1q/ osk2q/osk5q probes, respectively (Figure 5H).



CONCLUSION We introduced NIR-labeled DNA FIT probes−termed qFIT probes−to localize and quantify oskar mRNA in Drosophila melanogaster egg chambers at stage 8−9, when oskar mRNA accumulates in the posterior domain (PD) of the oocyte. We found thatas reported previously52a mean 55% of the total oskar mRNA localized to the 3000 μm3 large volume at the posterior pole that corresponds to ∼2.4% of the total oocyte volume. The measurements revealed a mean concentration of 200 nM of oskar within the PD. In 0.075 μm3 volume units, the enrichment can lead to a remarkably high local concentration of 1.8 μM. Surprisingly, we found that, while the local concentration of oskar was increasing continuously, the total number of oskar copies was independent of the oocyte size during stages 8 and 9 of oogenesis. As transcription of oskar mRNA does not appear to cease in the nurse cells during this phase of oogenesis, intraooplasmic levels of oskar mRNA must be regulated on a post-transcriptional level, e.g., on the level of RNA transport and/or RNA turnover. Also, our data indicate that this regulatory mechanism is not triggered by reaching a certain RNA copy number or RNA concentration, as we detected roughly half as much oskar in the oocytes of egg chambers that lack one functional allele of the oskar locus (Figure 5F). The qFIT probes are oligonucleotide-based hybridization probes that combine an internally positioned thiazole orange (TO) dye as a surrogate nucleobase, known from FIT probes, with a terminally appended near-infrared (NIR) dye. We demonstrated that the Cy7 dye is a suitable concentration reporter and showed that the brightness of TO emission remained high requiring that the Cy7-TO distance was sufficiently large (≥18 nucleotides) to avoid fluorescence resonance energy transfer (FRET) in the double-stranded state. Of note, the feasibility of FRET in the single-stranded state increased the fold-change (from up to 9-fold for the TOonly probes to up to 17-fold for Cy7/TO probes) of TO emission upon hybridization. To test the accuracy of the qFIT probe based assay, we compared oskar mRNA copy numbers determined by our assay and by single molecule FISH (smFISH).13 We found very strong linear, almost one-to-one correlation between the copy numbers reported by the two assays (Figure S9). The fluorescence microscopy imaging of oskar mRNA in Drosophila oocytes using a single qFIT probe at a threshold level that leads to 25% false positive detection rate revealed a detection limit of five copies within a 0.075 μm−3 volume unit of the fixed specimen. Higher threshold values allowed for a substantial decrease of the FPDR down to 0%. At this absolute confidence level, the limit of detection was increased to 12 copies per 0.075 μm−3 volume unit. Obviously, this sensitivity is not sufficient for the analysis of single molecules. However, the



METHODS



ASSOCIATED CONTENT

Clicking Cy7 onto FIT-DNA. To butyne-1-ol-modified FIT probes (100 μM) in buffer (100 mM TRIS, 100 mM NaOAc, 1 mM MgCl2, pH 8) was added ascorbate (7.5 mM), THPTA (3.75 mM), and the Cy-azide 1 (500 μM) in DMSO. After the addition of copper(II) sulfate (0.75 mM), the reaction mixture was agitated for 3 h at 55 °C. The precipitate obtained upon addition of iPrOH was collected by centrifugation. The pellet was washed twice with iPrOH and dissolved in water (500 μL). Insoluble material was removed by means of a syringe filter (0.45 μm). The oligonucleotides were purified by RPHPLC by using an X-Bridge BEH130 C18-column and a linear gradient of a binary solvent mixture 10% to 70% B within 12 min (A: 0.1 M triethylammonium acetate, aq. pH 7.5; B: MeCN) at a flow rate of 1.5 mL/min. The HPLC-purified compounds were submitted to a second iPrOH precipitation. Wash-Free in Situ Hybridization. Ovaries were fixed in 4% PFA for 20 min at ambient temperature. After two short washes with IBEX (10 mM HEPES, p = 7.7, 100 mM KCl, 1 mM EDTA, 0.3% Triton-X100), the ovaries were subjected to 92 °C for 3 min in preheated IBEX. The temperature was rapidly reduced to RT by adding 2× volume of ice-cold IBEX and then the hybridization solution prewarmed to 37 °Cwas administered. The hybridization solution was IBEX supplemented with 15% ethylene carbonate (IBEX-EC) and 1 w/v% SDS. After 20 min incubation, the hybridization solution was replaced with a fresh solution. The ovaries in 2−3 μL hybridization buffer were then mounted on a slide under Voltalef 10S halocarbon oil that prevents evaporation of water from the specimen. A coverslip was placed on the preparation, and by gentle pressure, the oocytes were slightly squeezed between the two glass surfaces to reduce their thickness and facilitate subsequent analyses. Microscopy. Samples were imaged with a Leica SP8 confocal microscope using a 60× water immersion objective (NA = 1.2). TO and Cy7 was excited with a 514 and 660 nm laser line produced by a white-light laser light source, and their emission was detected in 525− 570 nm and 700−800 nm, respectively.

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acschembio.7b01007. Materials and methods, characterization data, and additional results (PDF) 747

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology



Translation by Visualizing Ribosome−mRNA Interactions in Single Cells. ACS Cent. Sci. 3, 425−433. (16) Bao, G., Rhee, W. J., and Tsourkas, A. (2009) Fluorescent Probes for Live-Cell RNA Detection. Annu. Rev. Biomed. Eng. 11, 25− 47. (17) Tyagi, S. (2009) Imaging intracellular RNA distribution and dynamics in living cells. Nat. Methods 6, 331−338. (18) Tyagi, S., and Kramer, F. R. (1996) Molecular beacons: Probes that fluoresce upon hybridization. Nat. Biotechnol. 14, 303−308. (19) Tan, W. H., Wang, K. M., and Drake, T. J. (2004) Molecular beacons. Curr. Opin. Chem. Biol. 8, 547−553. (20) Wang, K. M., Tang, Z. W., Yang, C. Y. J., Kim, Y. M., Fang, X. H., Li, W., Wu, Y. R., Medley, C. D., Cao, Z. H., Li, J., Colon, P., Lin, H., and Tan, W. H. (2009) Molecular Engineering of DNA: Molecular Beacons. Angew. Chem., Int. Ed. 48, 856−870. (21) Köhler, O., Jarikote, D. V., and Seitz, O. (2005) Forced intercalation probes (FIT probes): Thiazole orange as a fluorescent base in peptide nucleic acids for homogeneous single-nucleotidepolymorphism detection. ChemBioChem 6, 69−77. (22) Hövelmann, F., Bethge, L., and Seitz, O. (2012) Single Labeled DNA FIT Probes for Avoiding False-Positive Signaling in the Detection of DNA/RNA in qPCR or Cell Media. ChemBioChem 13, 2072−2081. (23) Hövelmann, F., and Seitz, O. (2016) DNA Stains as Surrogate Nucleobases in Fluorogenic Hybridization Probes. Acc. Chem. Res. 49, 714−723. (24) Bethge, L., Singh, I., and Seitz, O. (2010) Designed thiazole orange nucleotides for the synthesis of single labelled oligonucleotides that fluoresce upon matched hybridization. Org. Biomol. Chem. 8, 2439−2448. (25) Kolpashchikov, D. M. (2010) Binary Probes for Nucleic Acid Analysis. Chem. Rev. 110, 4709−4723. (26) Marti, A. A., Jockusch, S., Stevens, N., Ju, J. Y., and Turro, N. J. (2007) Fluorescent hybridization probes for sensitive and selective DNA and RNA detection. Acc. Chem. Res. 40, 402−409. (27) Cardullo, R. A., Agrawal, S., Flores, C., Zamecnik, P. C., and Wolf, D. E. (1988) Detection of nucleic acid hybridization by nonradiative fluorescence resonance ernergy transfer. Proc. Natl. Acad. Sci. U. S. A. 85, 8790−8794. (28) Okamoto, A. (2011) ECHO probes: a concept of fluorescence control for practical nucleic acid sensing. Chem. Soc. Rev. 40, 5815− 5828. (29) Ikeda, S., Kubota, T., Yuki, M., and Okamoto, A. (2009) Exciton-Controlled Hybridization-Sensitive Fluorescent Probes: Multicolor Detection of Nucleic Acids. Angew. Chem., Int. Ed. 48, 6480− 6484. (30) Abe, H., and Kool, E. T. (2006) Flow cytometric detection of specific RNAs in native human cells with quenched autoligating FRET probes. Proc. Natl. Acad. Sci. U. S. A. 103, 263−268. (31) Gorska, K., and Winssinger, N. (2013) Reactions Templated by Nucleic Acids: More Ways to Translate Oligonucleotide-Based Instructions into Emerging Function. Angew. Chem., Int. Ed. 52, 6820−6843. (32) Pianowski, Z., Gorska, K., Oswald, L., Merten, C. A., and Winssinger, N. (2009) Imaging of mRNA in Live Cells Using Nucleic Acid-Templated Reduction of Azidorhodamine Probes. J. Am. Chem. Soc. 131, 6492−6497. (33) Silverman, A. P., and Kool, E. T. (2005) Quenched probes for highly specific detection of cellular RNAs. Trends Biotechnol. 23, 225− 230. (34) Wu, H. X., Alexander, S. C., Jin, S. J., and Devaraj, N. K. (2016) A Bioorthogonal Near-Infrared Fluorogenic Probe for mRNA Detection. J. Am. Chem. Soc. 138, 11429−11432. (35) Chen, A. K., Behlke, M. A., and Tsourkas, A. (2007) Avoiding false-positive signals with nuclease-vulnerable molecular beacons in single living cells. Nucleic Acids Res. 35, e105. (36) Chen, A. K., Davydenko, O., Behlke, M. A., and Tsourkas, A. (2010) Ratiometric bimolecular beacons for the sensitive detection of RNA in single living cells. Nucleic Acids Res. 38, e148.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Oliver Seitz: 0000-0003-0611-4810 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS F.H. and O.S. are grateful for financial support from the Deutsche Forschungsgemeinschaft, Se 819-2/2 and SPP 1784. I.G. and A.E. are grateful for financial support from the European Molecular Biology Laboratory and the Deutsche Forschungsgemeinschaft, SPP 2333. We thank the EMBL Advanced Light Microscopy Facility and Leica for providing cutting-edge microscopy.



REFERENCES

(1) Mortazavi, A., Williams, B. A., McCue, K., Schaeffer, L., and Wold, B. (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat. Methods 5, 621−628. (2) Alwine, J. C., Kemp, D. J., and Stark, G. R. (1977) Method for the detection of specific RNAs in agarose gels by transfer to diazobenzyloxymethyl-paper and hybrdidization with DNA probes. Proc. Natl. Acad. Sci. U. S. A. 74, 5350−5354. (3) Chiang, P. W., Song, W. J., Wu, K. Y., Korenberg, J. R., Fogel, E. J., VanKeuren, M. L., Lashkari, D., and Kurnit, D. M. (1996) Use of a fluorescent-PCR reaction to detect genomic sequence copy number and transcriptional abundance. Genome Res. 6, 1013−1026. (4) Gibson, U. E. M., Heid, C. A., and Williams, P. M. (1996) A novel method for real time quantitative RT PCR. Genome Res. 6, 995− 1001. (5) Bertrand, E., Chartrand, P., Schaefer, M., Shenoy, S. M., Singer, R. H., and Long, R. M. (1998) Localization of ASH1 mRNA particles in living yeast. Mol. Cell 2, 437−445. (6) Daigle, N., and Ellenberg, J. (2007) lambda(N)-GFP: An RNA reporter system for live-cell imaging. Nat. Methods 4, 633−636. (7) Paige, J. S., Wu, K. Y., and Jaffrey, S. R. (2011) RNA Mimics of Green Fluorescent Protein. Science 333, 642−646. (8) Strack, R. L., Song, W. J., and Jaffrey, S. R. (2013) Using Spinachbased sensors for fluorescence imaging of intracellular metabolites and proteins in living bacteria. Nat. Protoc. 9, 146−155. (9) Sunbul, M., and Jaschke, A. (2013) Contact-Mediated Quenching for RNA Imaging in Bacteria with a Fluorophore-Binding Aptamer. Angew. Chem., Int. Ed. 52, 13401−13404. (10) Ying, Z. M., Wu, Z., Tu, B., Tan, W. H., and Jiang, J. H. (2017) Genetically Encoded Fluorescent RNA Sensor for Ratiometric Imaging of MicroRNA in Living Tumor Cells. J. Am. Chem. Soc. 139, 9779− 9782. (11) Ozawa, T., Natori, Y., Sato, M., and Umezawa, Y. (2007) Imaging dynamics of endogenous mitochondrial RNA in single living cells. Nat. Methods 4, 413−419. (12) Nelles, D. A., Fang, M. Y., O’Connell, M. R., Xu, J. L., Markmiller, S. J., Doudna, J. A., and Yeo, G. W. (2016) Programmable RNA Tracking in Live Cells with CRISPR/Cas9. Cell 165, 488−496. (13) Raj, A., van den Bogaard, P., Rifkin, S. A., van Oudenaarden, A., and Tyagi, S. (2008) Imaging individual mRNA molecules using multiple singly labeled probes. Nat. Methods 5, 877−879. (14) Shah, S., Lubeck, E., Schwarzkopf, M., He, T., Greenbaum, A., Sohn, C. H., Lignell, A., Choi, H. M. T., Gradinaru, V., Pierce, N. A., and Cai, L. (2016) Single-Molecule RNA Detection at Depth via Hybridization Chain Reaction and Tissue Hydrogel Embedding and Clearing. Development 143, 2862−2867. (15) Burke, K. S., Antilla, K. A., and Tirrell, D. A. (2017) A Fluorescence in Situ Hybridization Method To Quantify mRNA 748

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749

Articles

ACS Chemical Biology (37) Zhang, X. M., Song, Y., Shah, A. Y., Lekova, V., Raj, A., Huang, L., Behlke, M. A., and Tsourkas, A. (2013) Quantitative assessment of ratiometric bimolecular beacons as a tool for imaging single engineered RNA transcripts and measuring gene expression in living cells. Nucleic Acids Res. 41, e152. (38) Zhang, X. M., Zajac, A. L., Huang, L. Y., Behlke, M. A., and Tsourkas, A. (2014) Imaging the Directed Transport of Single Engineered RNA Transcripts in Real-Time Using Ratiometric Bimolecular Beacons. PLoS One 9, e85813. (39) Seitz, O., Bergmann, F., and Heindl, D. (1999) A convergent strategy for the modification of peptide nucleic acids: Novel mismatchspecific PNA-hybridization probes. Angew. Chem., Int. Ed. 38, 2203− 2206. (40) Hövelmann, F., Gaspar, I., Chamiolo, J., Kasper, M., Steffen, J., Ephrussi, A., and Seitz, O. (2016) LNA-enhanced DNA FIT-probes for multicolour RNA imaging. Chem. Sci. 7, 128−135. (41) Hövelmann, F., Gaspar, I., Ephrussi, A., and Seitz, O. (2013) Brightness Enhanced DNA FIT-Probes for Wash-Free RNA Imaging in Tissue. J. Am. Chem. Soc. 135, 19025−19032. (42) Hövelmann, F., Gaspar, I., Loibl, S., Ermilov, E. A., Roder, B., Wengel, J., Ephrussi, A., and Seitz, O. (2014) Brightness through Local Constraint-LNA-Enhanced FIT Hybridization Probes for In Vivo Ribonucleotide Particle Tracking. Angew. Chem., Int. Ed. 53, 11370− 11375. (43) Kummer, S., Knoll, A., Socher, E., Bethge, L., Herrmann, A., and Seitz, O. (2011) Fluorescence Imaging of Influenza H1N1 mRNA in Living Infected Cells Using Single-Chromophore FIT-PNA. Angew. Chem., Int. Ed. 50, 1931−1934. (44) Kummer, S., Knoll, A., Socher, E., Bethge, L., Herrmann, A., and Seitz, O. (2012) PNA FIT-Probes for the Dual Color Imaging of Two Viral mRNA Targets in Influenza H1N1 Infected Live Cells. Bioconjugate Chem. 23, 2051−2060. (45) Karunakaran, V., Perez Lustres, J. L., Zhao, L. J., Ernsting, N. P., and Seitz, O. (2006) Large dynamic stokes shift of DNA intercalation dye thiazole orange has contribution from a high-frequency mode. J. Am. Chem. Soc. 128, 2954−2962. (46) Socher, E., Knoll, A., and Seitz, O. (2012) Dual fluorophore PNA FIT-probes - extremely responsive and bright hybridization probes for the sensitive detection of DNA and RNA. Org. Biomol. Chem. 10, 7363−7371. (47) Tornøe, C. W., Christensen, C., and Meldal, M. (2002) Peptidotriazoles on solid phase: 1,2,3 -triazoles by regiospecific copper(I)-catalyzed 1,3-dipolar cycloadditions of terminal alkynes to azides. J. Org. Chem. 67, 3057−3064. (48) Rostovtsev, V. V., Green, L. G., Fokin, V. V., and Sharpless, K. B. (2002) A stepwise Huisgen cycloaddition process: Copper(I)catalyzed regioselective ″ligation″ of azides and terminal alkynes. Angew. Chem., Int. Ed. 41, 2596−2599. (49) Bratu, D. P., Cha, B. J., Mhlanga, M. M., Kramer, F. R., and Tyagi, S. (2003) Visualizing the distribution and transport of mRNAs in living cells. Proc. Natl. Acad. Sci. U. S. A. 100, 13308−13313. (50) El-Sagheer, A. H., and Brown, T. (2011) Efficient RNA synthesis by in vitro transcription of a triazole-modified DNA template. Chem. Commun. 47, 12057−12058. (51) Little, S. C., Sinsimer, K. S., Lee, J. J., Wieschaus, E. F., and Gavis, E. R. (2015) Independent and coordinate trafficking of single Drosophila germ plasm mRNAs. Nat. Cell Biol. 17, 558−568. (52) Gaspar, I., Yu, Y. X. V., Cotton, S. L., Kim, D. H., Ephrussi, A., and Welte, M. A. (2014) Klar ensures thermal robustness of oskar localization by restraining RNP motility. J. Cell Biol. 206, 199−215.

749

DOI: 10.1021/acschembio.7b01007 ACS Chem. Biol. 2018, 13, 742−749